linux/mm
Damien Ramonda af248a0c67 readahead: fix sequential read cache miss detection
The kernel's readahead algorithm sometimes interprets random read
accesses as sequential and triggers unnecessary data prefecthing from
storage device (impacting random read average latency).

In order to identify sequential cache read misses, the readahead
algorithm intends to check whether offset - previous offset == 1
(trivial sequential reads) or offset - previous offset == 0 (sequential
reads not aligned on page boundary):

  if (offset - (ra->prev_pos >> PAGE_CACHE_SHIFT) <= 1UL)

The current offset is stored in the "offset" variable of type "pgoff_t"
(unsigned long), while previous offset is stored in "ra->prev_pos" of
type "loff_t" (long long).  Therefore, operands of the if statement are
implicitly converted to type long long.  Consequently, when previous
offset > current offset (which happens on random pattern), the if
condition is true and access is wrongly interpeted as sequential.  An
unnecessary data prefetching is triggered, impacting the average random
read latency.

Storing the previous offset value in a "pgoff_t" variable (unsigned
long) fixes the sequential read detection logic.

Signed-off-by: Damien Ramonda <damien.ramonda@intel.com>
Reviewed-by: Fengguang Wu <fengguang.wu@intel.com>
Acked-by: Pierre Tardy <pierre.tardy@intel.com>
Acked-by: David Cohen <david.a.cohen@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-11-13 12:09:09 +09:00
..
backing-dev.c mm/backing-dev.c: check user buffer length before copying data to the related user buffer 2013-09-11 15:58:03 -07:00
balloon_compaction.c
bootmem.c mm/bootmem.c: remove unused local `map' 2013-11-13 12:09:09 +09:00
bounce.c mm/bounce.c: fix a regression where MS_SNAP_STABLE (stable pages snapshotting) was ignored 2013-09-30 14:31:02 -07:00
cleancache.c mm: cleancache: clean up cleancache_enabled 2013-04-30 17:04:01 -07:00
compaction.c mm/compaction.c: update comment about zone lock in isolate_freepages_block 2013-11-13 12:09:03 +09:00
debug-pagealloc.c
dmapool.c
fadvise.c teach SYSCALL_DEFINE<n> how to deal with long long/unsigned long long 2013-03-03 22:46:22 -05:00
failslab.c
filemap_xip.c lift sb_start_write() out of ->write() 2013-04-09 14:12:56 -04:00
filemap.c mm: memcg: handle non-error OOM situations more gracefully 2013-10-16 21:35:53 -07:00
fremap.c mm: save soft-dirty bits on file pages 2013-08-13 17:57:48 -07:00
frontswap.c frontswap: fix incorrect zeroing and allocation size for frontswap_map 2013-06-12 16:29:46 -07:00
highmem.c Some nice cleanups, and even a patch my wife did as a "live" demo for 2012-12-20 08:37:05 -08:00
huge_memory.c mm: thp: khugepaged: add policy for finding target node 2013-11-13 12:09:06 +09:00
hugetlb_cgroup.c cgroup: pass around cgroup_subsys_state instead of cgroup in file methods 2013-08-08 20:11:24 -04:00
hugetlb.c mm: hugetlb: initialize PG_reserved for tail pages of gigantic compound pages 2013-10-16 21:35:52 -07:00
hwpoison-inject.c mm/hwpoison: fix the lack of one reference count against poisoned page 2013-09-30 14:31:03 -07:00
init-mm.c
internal.h mm: vmscan: fix do_try_to_free_pages() livelock 2013-09-11 15:58:01 -07:00
interval_tree.c
Kconfig mem-hotplug: introduce movable_node boot option 2013-11-13 12:09:09 +09:00
Kconfig.debug
kmemcheck.c
kmemleak-test.c
kmemleak.c mm: kmemleak: avoid false negatives on vmalloc'ed objects 2013-11-13 12:09:07 +09:00
ksm.c ksm: remove redundant __GFP_ZERO from kcalloc 2013-11-13 12:09:02 +09:00
list_lru.c mm: list_lru: fix almost infinite loop causing effective livelock 2013-10-30 12:57:46 -07:00
maccess.c
madvise.c mm/hwpoison: fix traversal of hugetlbfs pages to avoid printk flood 2013-09-30 14:31:02 -07:00
Makefile list: add a new LRU list type 2013-09-10 18:56:30 -04:00
memblock.c mm/memblock.c: introduce bottom-up allocation mode 2013-11-13 12:09:08 +09:00
memcontrol.c memcg: support hierarchical memory.numa_stats 2013-11-13 12:09:06 +09:00
memory_hotplug.c mem-hotplug: introduce movable_node boot option 2013-11-13 12:09:09 +09:00
memory-failure.c mm/memory-failure.c: move set_migratetype_isolate() outside get_any_page() 2013-11-13 12:09:04 +09:00
memory.c mm: remove obsolete comments about page table lock 2013-11-13 12:09:03 +09:00
mempolicy.c mm/mempolicy: use NUMA_NO_NODE 2013-11-13 12:09:06 +09:00
mempool.c mm/mempool.c: convert kmalloc_node(...GFP_ZERO...) to kzalloc_node(...) 2013-09-11 15:58:14 -07:00
migrate.c Merge branch 'linus' into sched/core 2013-11-01 08:24:41 +01:00
mincore.c swap: make each swap partition have one address_space 2013-02-23 17:50:17 -08:00
mlock.c mm/mlock.c: prevent walking off the end of a pagetable in no-pmd configuration 2013-09-30 14:31:02 -07:00
mm_init.c mm: numa: Change page last {nid,pid} into {cpu,pid} 2013-10-09 14:47:45 +02:00
mmap.c mmap: arch_get_unmapped_area(): use proper mmap base for bottom up direction 2013-11-13 12:09:08 +09:00
mmu_context.c mm: remove old aio use_mm() comment 2013-05-07 18:38:27 -07:00
mmu_notifier.c treewide: relase -> release 2013-06-28 14:34:33 +02:00
mmzone.c mm: numa: Change page last {nid,pid} into {cpu,pid} 2013-10-09 14:47:45 +02:00
mprotect.c Merge branch 'linus' into sched/core 2013-11-01 08:24:41 +01:00
mremap.c mm: revert mremap pud_free anti-fix 2013-10-16 21:35:53 -07:00
msync.c
nobootmem.c mm/nobootmem.c: have __free_pages_memory() free in larger chunks. 2013-11-13 12:09:04 +09:00
nommu.c mm: remove free_area_cache 2013-07-10 18:11:34 -07:00
oom_kill.c mm: memcg: handle non-error OOM situations more gracefully 2013-10-16 21:35:53 -07:00
page_alloc.c mm/page_alloc.c: remove unused marco LONG_ALIGN 2013-11-13 12:09:07 +09:00
page_cgroup.c
page_io.c aio: Kill aio_rw_vect_retry() 2013-07-30 11:53:12 -04:00
page_isolation.c mm: memory-hotplug: enable memory hotplug to handle hugepage 2013-09-11 15:57:48 -07:00
page-writeback.c writeback: fix negative bdi max pause 2013-10-16 21:35:53 -07:00
pagewalk.c mm/pagewalk.c: fix walk_page_range() access of wrong PTEs 2013-10-30 14:27:03 -07:00
percpu-km.c
percpu-vm.c
percpu.c
pgtable-generic.c mm: move pgtable related functions to right place 2013-09-11 15:57:30 -07:00
process_vm_access.c Fix: compat_rw_copy_check_uvector() misuse in aio, readv, writev, and security keys 2013-03-12 11:05:45 -07:00
quicklist.c
readahead.c readahead: fix sequential read cache miss detection 2013-11-13 12:09:09 +09:00
rmap.c thp: account anon transparent huge pages into NR_ANON_PAGES 2013-09-12 15:38:03 -07:00
shmem.c initmpfs: make rootfs use tmpfs when CONFIG_TMPFS enabled 2013-09-11 15:59:37 -07:00
slab_common.c slab_common: Do not check for duplicate slab names 2013-09-28 09:47:41 +03:00
slab.c kernel: delete __cpuinit usage from all core kernel files 2013-07-14 19:36:59 -04:00
slab.h memcg: check that kmem_cache has memcg_params before accessing it 2013-08-28 19:26:38 -07:00
slob.c mm/sl[aou]b: Move kmallocXXX functions to common code 2013-09-04 20:51:33 +03:00
slub.c Merge branch 'slab/next' of git://git.kernel.org/pub/scm/linux/kernel/git/penberg/linux 2013-09-15 07:15:06 -04:00
sparse-vmemmap.c sparse-vmemmap: specify vmemmap population range in bytes 2013-04-29 15:54:35 -07:00
sparse.c mm/sparsemem: fix a bug in free_map_bootmem when CONFIG_SPARSEMEM_VMEMMAP 2013-11-13 12:09:06 +09:00
swap_state.c lib/radix-tree.c: make radix_tree_node_alloc() work correctly within interrupt 2013-09-11 15:59:36 -07:00
swap.c mm: make lru_add_drain_all() selective 2013-09-12 15:38:02 -07:00
swapfile.c frontswap: enable call to invalidate area on swapoff 2013-11-13 12:09:07 +09:00
truncate.c truncate: drop 'oldsize' truncate_pagecache() parameter 2013-09-12 15:38:02 -07:00
util.c swap: clean-up #ifdef in page_mapping() 2013-09-11 15:57:31 -07:00
vmalloc.c mm: kmemleak: avoid false negatives on vmalloc'ed objects 2013-11-13 12:09:07 +09:00
vmpressure.c Merge branch 'for-3.12' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup 2013-09-03 18:25:03 -07:00
vmscan.c mm/vmscan.c: don't forget to free shrinker->nr_deferred 2013-10-16 21:35:52 -07:00
vmstat.c mm: clear N_CPU from node_states at CPU offline 2013-11-13 12:09:09 +09:00
zbud.c mm/zbud: fix some trivial typos in comments 2013-09-11 15:57:35 -07:00
zswap.c mm/zswap: avoid unnecessary page scanning 2013-11-13 12:09:08 +09:00